SYNTAX: notion invoked in the analysis of preposition stranding and clause union phenomena. In the case of a pseudo-passive like (i)a the stranded preposition at and the verb laughed are sometimes reanalyzed as the complex verb laughed at as in (i)b.
(i) a Johni was laughed [PP at ti ] b Johni was [V laughed at] tiReanalysis eliminates the PP-boundary so that the NP-trace ti in (i)b can be properly governed by laughed. Another kind of reanalysis has been proposed in order to handle operations on thematic relations. In these cases reanalysis may collapse (or merge) two external theta-roles in restructuring constructions containing a modal verb, see (ii), but it may also demote an external theta-role to become an internal one in causative constructions, see (iii), while creating a complex predicate. Both examples are Italian:
(ii) Giorgio vuole sapere per che. Giorgio wants know why (iii) Giorgio fa dormire i bambini Giorgio makes sleep the childrenIn (ii) Giorgio would fulfill the external theta-roles of vuole and sapere which are merged to one. In (iii) i bambini fulfills the external theta-role of dormire, but this role would be turned into an internal role of the complex fa dormire. Since these kinds of operations are at odds with the theta criterion and the Projection Principle, two-dimensional phrase-structures have been proposed, assigning more than one representation simultaneously to one sentence, in order to circumvent these problems. One structure is determined basically by theta-relations corresponding to a non-reanalyzed D-structure, the other represents the reanalyzed clause union structure. See co-analysis.
|LIT.||Di Sciullo, A. M. and E. Williams (1987)|
Haegeman, L. & H. v.Riemsdijk (1986)
Koster, J. (1987)
Lasnik, H. & J. Kupin (1976)
Manzini, R. (1983a)
Zubizaretta, M.-L. (1987)
Zubizaretta, M.-L. (1982)